January 2017 Summaries

Migrating relational data from Oracle RDBMS to Neo4j may be necessary when an existing application reaches its performance limits, requiring a new application or significant development. This can occur due to slow application performance, complex queries, or the need for real-time joining of varying sets and depths of data. In such cases, migrating all data to Neo4j can greatly improve query times and simplify connected data queries. Companies like Schleich GmbH and Billes have successfully migrated their traditional systems to Neo4j, achieving improved flexibility, performance, and operational efficiency. The process involves editing application code, rewriting queries, and leveraging Neo4j's flexible data model for seamless system integration.

Graph databases like Neo4j provide significant advantages over relational databases when working with interconnected data, offering tools to uncover previously hidden relationships. A Ph.D. candidate in computer science uses Neo4j in her research on the impact assessment of schema evolution in a data warehouse context, finding it excels at capturing relationships and has a strong developer community and many new features related to security, clustering, scaling, and enterprise. Her favorite feature is graph visualization, which allows for easy path and relationship analysis, and she also uses Neo4j's REST API with Java drivers extensively in her research. She combines Neo4j with other technologies like relational schemas, Pentaho, and container orchestration, leveraging Neo4j's general representation of nodes, edges, and relationships to flatten heterogeneous artifacts. The candidate is happy to see Neo4j embracing Docker and its containerization capabilities, which have been incorporated into the database.

The text discusses the evolution of Neo4j, a popular graph database management system. In April 2016, Neo4j introduced its binary protocol named Bolt, which provided improved performance and efficiency for developers. The Bolt protocol was supported by various programming languages, including Java, Python, JavaScript, and .Net. Several contributors developed drivers for Bolt, including those in the Go, Elixir, and Haskell communities. In October 2016, Neo4j released its third beta version of Neo4j 3.1, which introduced new APIs to support Causal Clustering and Smart Client Routing. These features improved concurrent operations and session reuse, making it easier for developers to work with the graph database. The text also highlights the growing community of contributors who have developed Neo4j drivers, including those in the Ruby and PHP communities.

In a dental lab that manufactures prosthetic teeth, the Glidewell team uses Neo4j in various ways, including implementing a workflow engine into manufacturing processes and eventually all enterprise processes. They chose Neo4j for its expressive data representation and performance advantages, particularly with Cypher, which they found easy to use and incorporate into their systems. The team also uses MongoDB for catalogs and Elasticsearch for search systems, with most infrastructure built in C#-based microservices running in Docker containers. They have had success using Neo4j generically across the company's applications, leveraging its expressive relationships to create fundamental, generic applications that can be reused throughout the organization.

The text discusses the use of Neo4j, a graph database technology, to analyze and map the business connections of Donald J. Trump, as part of an investigative journalism project. The data was collected from a Google Spreadsheet and imported into Neo4j using Cypher queries. The resulting graph shows 1514 nodes (organizations) and 1857 relationships between them. The analysis reveals that organizations closely connected to Trump include Thrive Capital, 40 Wall Street LLC, and DJT Holdings LLC. The graph also shows the connections between people involved in Trump's cabinet nominations, including Ryan Zinke, Jeff Sessions, Wilbur Ross, Andy Puzder, Scott Pruitt, and Roma Downey. Additionally, the analysis explores transitive relationships within the data, finding shortest paths between individuals and organizations. The authors plan to extend the graph with related data from other sources, such as littlesis.org and OpenCorporates.

If you would like to see your post featured in January's "From the Community" blog post, follow us on Twitter and use the #Neo4j hashtag for your chance to get picked. Articles and blog posts covering various Neo4j topics are available, including running Neo4j 3.x on Docker, enterprise startup lists, and new releases. Podcasts and audio interviews with experts in the field are also featured, such as an interview with Emil Eifrem, Neo Technology's CEO. Videos showcasing graph database queries and various projects using Neo4j are available for viewing. Additionally, a list of libraries, GraphGists, and code repositories related to Neo4j is provided, including drivers and sample applications. The official Neo4j Certification can be completed in under an hour and showcases one's graph database skills to employers.

Neo4j and Oracle RDBMS can be fully synchronized to deliver better bottom-line results for enterprise architects and business teams. Applications that integrate data from multiple sources benefit from full synchronization, while existing applications writing to an Oracle database may require new technologies like Neo4j due to performance limitations. Monsanto, a multinational agrochemical company, successfully synchronized its Neo4j and Oracle Exadata databases using Apache Kafka as a distributed commit log to feed Neo4j with live transactional data. This synchronization enabled the company's genetic ancestry solution services to handle over 600 million REST requests across approximately a billion nodes, resulting in query times of just tenths of milliseconds.

Neo4j's graph query language, Cypher, has simplified the process of encoding and querying biological and medical knowledge in a network, reducing the lines of code from 1,000 to just 20. Daniel Himmelstein, a researcher at UPenn, uses Neo4j to represent complex relationships in real-world data, leveraging its community-driven features and support. He praises the ease of use, with Cypher providing an intuitive query language that has made it easier for him to work with Python and R within Neo4j. With the growing field of bioinformatics research expected to continue expanding, Himmelstein anticipates significant growth in this area over the next year.

Pavel Yakovlev, a director at BIOCAD, discusses the development of Hasbolt, a new Haskell driver for Neo4j's Bolt protocol, aimed at enhancing graph database application in biotechnology research. This initiative arises from the need for efficient data analysis tools in R&D for anti-cancer drugs, necessitating the use of graph databases due to their capacity to handle complex data relationships. Leveraging Haskell's functional programming paradigms like strong typing and immutable data structures, Yakovlev created a driver that supports serialization, deserialization, and network communication to facilitate interaction with Neo4j. The Hasbolt driver enables users to execute Cypher queries and manage connections within a Haskell environment, with the potential for integration into larger applications, such as a movie database example that illustrates various functionalities. This development is part of a broader exploration into graph databases for predictive analytics in biotechnology, highlighting the intersection of computational biology and advanced data management technologies.